Multi-logistics robot path planning method based on improved time window algorithm

Abstract

The present invention relates to a multi-logistics robot path planning method based on an improved time window algorithm, and belongs to the field of robot path planning technology. The method allocates handling tasks to each logistics robot in order according to task priority and robot priority; the robot is allocated according to task priority, and then the path planning is completed according to the starting coordinates and target coordinates of the handling task using the ant colony algorithm, and then the time window of the logistics robot on each path segment is calculated using the improved time window algorithm, and the time windows are arranged in order on each path, and the paths with time window conflicts are detected, and the paths are adjusted or replanned according to the time window conflict avoidance strategy until all paths do not have time window conflicts. The present invention can effectively solve the path conflict problem between multiple logistics robots in logistics scenarios, and ensure that multiple logistics robots can complete the predetermined handling tasks more safely and efficiently.

Description

Technical Field

[0001] The invention belongs to the technical field of robot path planning, and relates to a multi-logistics robot path planning method based on an improved time window algorithm. Background Art

[0002] The path planning problem of multiple robots is as follows: there are multiple logistics robots in the indoor warehouse waiting for the assignment of handling tasks in the waiting area. Several logistics robots get the handling tasks, and then use the set global path planning algorithm to complete the path planning for each robot. Since multiple logistics robots will run at the same time and the operating environment is limited, the running paths of each logistics robot will have path conflicts at some nodes. Therefore, it is necessary to consider which logistics robot will pass first and which robot will give way at the conflicting node to ensure that each logistics robot completes the handling task without collision. This requires considering the planning priority of each logistics robot and the priority of the handling task. The path planning of multiple logistics robots is to reasonably allocate the running time and space of the robots, complete the coordinated operation of all logistics robots, and ensure that each logistics robot completes the handling task in an orderly and collision-free manner.

[0003] The commonly used path planning based on the time window algorithm is to calculate the occupied time period of each logistics robot's path on each path, including the entry time and exit time of the logistics robot on the grid, arrange the time windows of each logistics robot in the order of the path, and then detect whether there is only one logistics robot passing through each path in the same time period. If there are multiple logistics robots that need to pass through the same time period on the path, there will be path conflicts between multiple logistics robots. By analyzing the time window occupancy on each path, the time windows are rearranged according to the task priority of each logistics robot, so that only one logistics robot passes through at the same time on any path. The computational complexity of the time window algorithm will increase with the increase of the grid map scale and the number of logistics robots.

[0004] When a time window conflict is detected on a path, the time window conflict is often resolved through three processing measures: translation, stretching, and compression of the time window. The translation operation is to make the low-priority robot stop and wait between the conflicting nodes or slow down until the high-priority robot passes the conflicting node, stretch the conflicting time window of the low-priority robot, and translate the subsequent time windows backward in chronological order. Finally, it is ensured that there is no conflict in the time windows on each path. The compression operation requires the robot to accelerate to shorten the running time on a certain path. This operation will increase the calculation amount of the algorithm and increase the volatility of the operation.

[0005] In actual use in warehousing environments, when multiple logistics robots operate in collaboration, there are still problems such as path occupancy and path conflicts caused by the operation of the logistics robots.

[0006] Therefore, there is an urgent need for a multi-logistics robot path planning method that can simultaneously solve the path conflict problem and the time window algorithm problem in the existing methods. Summary of the invention

[0007] In view of this, the purpose of the present invention is to provide a multi-logistics robot path planning method based on an improved time window algorithm, which effectively solves the path conflict problem between multiple logistics robots in logistics scenarios and ensures that multiple logistics robots can complete the predetermined handling tasks more safely and efficiently.

[0008] In order to achieve the above object, the present invention provides the following technical solutions:

[0009] A multi-logistics robot path planning method based on an improved time window algorithm, the method orderly assigns handling tasks to each logistics robot according to task priority and robot priority; assigns high-priority handling tasks to high-priority robots, and assigns low-priority handling tasks to low-priority logistics robots, then uses an ant colony algorithm to complete path planning according to the starting coordinates and target coordinates of the handling tasks, and then uses an improved time window algorithm to calculate the time windows of the logistics robots on each path segment, arranges the time windows in order on each path, detects paths with time window conflicts, and adjusts or replans the paths according to time window conflict avoidance strategies until all paths have no time window conflicts.

[0010] The method specifically comprises the following steps:

[0011] S1: Assign tasks to logistics robots in order according to their priority;

[0012] S2: Use the ant colony algorithm to complete the path planning for the logistics robot, and complete the time window calculation of each logistics robot according to the improved time window algorithm;

[0013] S3: Determine whether the time windows of the current highest priority logistics robot and the second highest priority logistics robot overlap. If not, jump to step 7; if there is a time window overlap, execute step S4;

[0014] S4: Determine whether the path conflict type with time overlap is a node conflict. If so, the low-priority logistics robot stops and waits at the node before the conflicting node until the high-priority logistics robot passes the node and completes the time window update; otherwise, execute step S5;

[0015] S5: Determine whether the path conflict type with time overlap is an opposite conflict. If it is an opposite conflict, calculate the parking waiting time of the low-priority logistics robot from the previous node of the conflicting path, remove the conflicting path from the search list, and re-plan the running time required for the path to the target point with the node as the starting point. Finally, select the plan with the shortest time consumption, update the time window of the plan and jump to step S3; if there is no opposite conflict, execute step S6;

[0016] S6: Determine whether the path conflict type with time overlap is a chasing conflict. If it is a chasing conflict, perform road right analysis on the two logistics robots. If the driving order of the two will not cause obstacles for each other to reach the target point, jump to step S3; if it will cause obstacles, remove the conflicting path nodes of the two logistics robots from the search list, use the ant colony algorithm to search for a suboptimal path for the low-priority logistics robot again, complete the time window update, and jump to step S3;

[0017] S7: Determine whether all tasks are completed. If all tasks are completed, continue to step S8; if not, jump to step S3;

[0018] S8: End.

[0019] Furthermore, in step S1, the rules for assigning task priorities are as follows: (1) handling tasks are assigned in order of priority from high to low; (2) each logistics robot can only be assigned one task at a time; (3) each handling task can only be assigned to one logistics robot.

[0020] Furthermore, in step S1, each logistics robot has a fixed ID number, and the priority of the logistics robot is determined by the ID number. The robot with a high priority can get the handling task first, and the smaller the ID number, the higher the priority;

[0021] The logistics robot status is divided into: picking status, running status, idle status, charging status, and fault status; logistics robots in charging status and fault status will not appear in the list of handling task assignments; logistics robots in picking status indicate that they are waiting for goods to be picked in the shelf area, and they do not exist in the sequence of task assignments; robots in running status indicate that they have completed task assignments and are going to the picking area or have completed picking and are going to the target area to deliver goods; robots in idle status will appear in the robot list of task assignments, and the list ranking order is arranged according to the size of the ID number; the ID number of the logistics robot in idle status determines the priority of the robot. The smaller the ID number, the higher the priority, and the handling task is assigned first.

[0022] Furthermore, the improved time window algorithm includes: after considering the state of the logistics robot, if two logistics robots have the same priority, the priority of the logistics robot is determined by the relationship of the path occupancy right;

[0023] If the target point of logistics robot R1 is on the optimal path of logistics robot R2, then ρ1<ρ2 should be satisfied, and R2 has the right to prioritize path planning to avoid deadlock; where ρ1 is the path occupation right of R1, and ρ2 is the path occupation right of R2;

[0024] If the starting point of logistics robot R1 is on the optimal path of logistics robot R2, then ρ1>ρ2 should be satisfied, and R1 has the right to priority path planning to avoid deadlock.

[0025] When multiple logistics robots occupy each other's priorities, different conflict avoidance strategies are adopted according to different conflict types.

[0026] Furthermore, after completing the task allocation and path planning, the time window algorithm is used to perform time window detection on the task path to determine whether there is a time window conflict between the paths. If there is no conflict, it can be executed according to the task path. If there is a conflict, the corresponding conflict avoidance strategy is adopted according to the corresponding conflict type to complete the time window arrangement and resolve the time window conflict. The conflict avoidance strategies of the time window algorithm generally include parking waiting, path replanning, and speed adjustment. The corresponding methods are three kinds of operation methods: translation, stretching, and compression of the time window. The time window translation operation refers to the low-priority robot stopping in front of the conflict node and waiting for the high-priority robot to pass first. The low-priority robot passes the node again, and the time window of the low-priority robot needs to be stretched and translated backward. If the two operations of time window stretching and compression involve the acceleration and deceleration of the robot, it will lead to an increase in the amount of calculation and the uncontrollable factors of the system.

[0027] The two measures commonly used in the avoidance strategy of time window conflicts are parking and waiting and path replanning. When multiple robots work at the same time, the path situation is relatively complicated. If only the parking and waiting strategy is adopted, it will cause blockages on each path, affecting the working efficiency of the logistics system. If the path replanning strategy is adopted, the handling task cannot be completed, because the time window conflict avoidance is started according to the task priority. The high-priority robots occupy too many paths, and the subsequent robots will fall into the problem of having no path to go. In order to solve the problem of deadlock and blockage of path conflict nodes, the present invention designs a conflict avoidance strategy that combines parking and waiting with path replanning, and the path replanning is divided into primary replanning and secondary replanning.

[0028] The path planning order of the logistics robot is obtained through priority division, and different conflict avoidance strategies are adopted according to different conflict types. If there is a node conflict, the path conflict avoidance can be completed by just stopping and waiting. For chasing conflicts and opposite conflicts, the low-priority robot needs to replan the path to solve the conflict problem, especially when the paths between robots are still occupied by each other, it is more necessary to analyze the right of way and select the appropriate avoidance strategy. The first-level path replanning means that when a time window conflict is detected, the conflicting nodes are removed from the search list, and the ant colony algorithm is used to replan the path from the starting point to the target point for the low-priority robot. If there is no conflict in the time window of the replanned path, it means that the replanning is successful. If there is still a time window conflict, it means that the replanning fails and it is necessary to wait for the position of other robots to change before replanning the path. The second-level path replanning means that by traversing the path information, the previous node of the conflicting node is found, and the path from the node to the target point is replanned for the low-priority robot. The time cost of the parking waiting and replanning measures is compared, and the shortest time solution is selected for execution.

[0029] The beneficial effects of the present invention are as follows: the present invention is based on a local path planning algorithm based on an improved time window algorithm, and also comprehensively considers the path occupancy situation to reformulate the task priority and robot priority allocation rules, which can effectively solve the path conflict problem between multiple logistics robots in logistics scenarios, and ensure that multiple logistics robots can complete the predetermined handling tasks more safely and efficiently.

[0030] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0032] Figure 1 A schematic diagram of the driving route;

[0033] Figure 2 It is a schematic diagram of the time window in the basic time window algorithm;

[0034] Figure 3 This is a schematic diagram of path conflict;

[0035] Figure 4 To improve the time window diagram in the time window algorithm;

[0036] Figure 5This is a schematic diagram of the time window after path planning, where: Figure 5 (a) is the time window before extension, Figure 5 (b) is the time window after extension;

[0037] Figure 6 This is a flow chart of a multi-logistics robot path planning method based on an improved time window algorithm of the present invention;

[0038] Figure 7 It is a grid map of the warehouse logistics operation environment;

[0039] Figure 8 It is a path planning diagram based on the improved ant colony algorithm;

[0040] Fig. 9 It is a schematic diagram of the time windows of R1 and R2;

[0041] Fig.10 This is a schematic diagram of the time windows of R1 and R2 after planning;

[0042] Fig.11 It is a schematic diagram of the time windows of R3 and R4;

[0043] Fig.12 This is a schematic diagram of the time window of R3 and R4 after planning;

[0044] Fig.13 This is the path planned based on the improved time window algorithm. DETAILED DESCRIPTION

[0045] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0046] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on the present invention. In order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0047] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "front", "rear", etc. indicate the orientation or position relationship, they are based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0048] See also Figures 1 to 13 The present invention provides a multi-logistics robot path planning method based on an improved time window algorithm, and performs local path planning based on the improved time window algorithm. The improved world time window algorithm is improved in terms of priority allocation method and time window conflict avoidance strategy.

[0049] 1. Improvement of priority allocation method

[0050] The time window algorithm usually uses two methods: assigning task priority and logistics robot priority, which can ensure that the cargo handling task is handled and completed in a timely and effective manner. The logistics system assigns priorities to tasks in advance when assigning tasks. The higher the priority of the handling task, the priority of the task will be assigned. If a path conflict occurs when the logistics robot is performing the task, the order of path planning can be determined according to the priority of the task. When a path conflict or mutual obstacle avoidance occurs, the logistics robot performing the high-priority task will maintain normal operation, and the robot performing the low-priority task will take measures such as avoidance or stopping operation to ensure that each logistics robot runs without collision. The following rules are followed during the task priority assignment process:

[0051] (1) The handling tasks are assigned in order of priority from high to low;

[0052] (2) Each logistics robot can only be assigned one task at a time;

[0053] (3) Each handling task can only be assigned to one logistics robot.

[0054] Each logistics robot has a fixed ID number. The priority of the logistics robot is determined by the ID number. The robot with a high priority can get the handling task first. The smaller the ID number, the higher the priority. The status of the logistics robot is divided into: picking state, running state, idle state, charging state, and fault state. The logistics robot in the charging state and fault state will not appear in the list of handling task assignment; the logistics robot in the picking state indicates that it is waiting for goods to be picked in the shelf area, and it does not exist in the sequence of task assignment; the robot in the running state indicates that the task assignment has been completed and it is going to the picking area or has completed picking and is going to the target area to deliver goods; the robot in the idle state will appear in the robot list of task assignment, and the ranking order of the list is arranged according to the size of the ID number. The ID number of the logistics robot in the idle state determines the priority of the robot. The smaller the ID number, the higher the priority, and the handling task is assigned first.

[0055] Based on the basic time window algorithm, the present invention obtains the local path planning order of the robot according to the task priority. If logistics robots with the same priority appear, the planning order is considered according to the distance between the robot and the conflict grid. In the planning process, the robot can deviate from the optimal path for avoidance. After the high-priority robot passes the conflict node first, the low-priority robot re-plans the path to the target location, thereby improving the success probability of local path planning.

[0056] The basic time window algorithm is: Assume that there are n logistics robots in the logistics system waiting for task allocation and running collaboratively at the same time, and the set of logistics robots is R = {R1, R2, R3, R4...R n Among them, there are m logistics robots assigned with transportation tasks, and the task set is M = {M1, M2, M3, M4...M m}, each task M i All have priorities. After completing the handling task allocation, the global path planning algorithm is used to complete the optimal path planning for each logistics robot. The path segment set of each logistics robot is L = {l1,l2,l3,l4......l k First, the path is divided into path segments according to the grid, and the time window is initialized for the grids contained in the path. i The first grid (the current location of the logistics robot), assuming that R1 enters the path l1 grid g i The time is (1 is the number of the logistics robot, i is the grid number, and in is the time of entering the grid), leaving l1 and arriving at grid g j The time is Then define the path l1 grid g i → j The time window reserved by R1 is It indicates the time that R1 occupies in the path segment l1. After the time window initialization operation, the time window of each logistics robot path on each path segment can be obtained. The time window set is Then, the time windows are inserted into each path segment in order. The sorted time window set of R1 is T1 = {t 1,1 ,t 1,2 ,t 1,3 ,…,t 1,n}, the entry time set of R1 is expressed as The departure time set of R1 is expressed as Since each grid or each path is only allowed to be occupied by one logistics robot at a time, the time windows of each grid or each path should not overlap, and the mathematical relationship of the time windows is:

[0057]

[0058] Taking the handling task of two logistics robots as an example, the calculation process of the time window mathematical model is discussed in detail. The working environment map of the logistics robot is represented by M = {g1, g2, g3, g4, …, g n}, where g i Indicates the grid number. The schematic diagram of the logistics robot's driving path is as follows Figure 1 As shown in the figure, the starting position of R1 is grid g1, the target point is grid g8, the starting position of R2 is grid g6, and the target point is grid g7. The ant colony algorithm is used to plan the paths for the two logistics robots. The optimal path of R1 is g1→g2→g3→g8, and the optimal path of R2 is g6→g5→g4→g3→g7.

[0059] After completing the global path planning, the logistics robot starts to drive along the optimal path. The start and braking time of the logistics robot are set to t, the driving speed is constant at V, and the steering speed is constant at V t , then the travel time of the logistics robot between each grid can be calculated. Assume that R1 enters g i The time is Leave the grid j The time is Then the time window for a certain path segment to be reserved by R1 is The set of time windows for all reservations is represented as According to the path information and the driving speed of the logistics robot, the occupancy time on each path can be calculated. The residence time between R1 grids is shown in Table 1.

[0060] Table 1 R1 grid dwell time

[0061]

[0062] Similarly, the stay time of R2 between each path segment can be obtained, as shown in Table 2.

[0063] Table 2 R2 grid dwell time

[0064]

[0065] Sort the time windows of each logistics robot on each path in order of the paths. The results are as follows: Figure 2 shown.

[0066] The time window set of R1 after sorting is The entry time is expressed as The departure time is expressed as The mathematical relationship of the R1 time window is shown in formula (2). The time window set of R2 after sorting is The entry time is expressed as The departure time is expressed as The mathematical relationship of the R2 time window is shown in formula (3). Since each grid or each path segment is only allowed to be occupied by one robot at a time, the time windows of each grid or each path segment should not overlap.

[0067]

[0068]

[0069] When a time window conflict is detected on a path, the time window conflict can be resolved through three processing measures: translation, stretching, and compression of the time window. The translation operation is to make the low-priority robot stop and wait between the conflicting nodes or slow down until the high-priority robot passes the conflicting node, stretch the conflicting time window of the low-priority robot, and translate the subsequent time windows backward in chronological order. Finally, it is ensured that there is no conflict in the time windows on each path. The compression operation requires the robot to accelerate to shorten the running time on a certain path. This operation will increase the calculation amount of the algorithm and increase the volatility of the operation.

[0070] An example of two logistics robots having a path conflict is Figure 3 As shown in the figure, the two logistics robots are assigned tasks, the target point of R1 is g5, and the target point of R2 is g6. The ant colony algorithm is used to complete the path planning for the two logistics robots, and the path of R1 is g1→g2→g3→g4→g5, and the path of R2 is g5→g4→g3→g6.

[0071] According to the path obtained by path planning, the occupied time of the two logistics robots on each path is calculated, and the time windows of the two logistics robots are sorted according to the path order, such as Figure 4 shown. Figure 4 It can be seen that the time window overlaps in grid g3, indicating that the two logistics robots will have a path conflict. The two robots will collide with each other. Since the robots are equipped with obstacle avoidance sensors, although the two logistics robots will not collide with each other, they will fall into a deadlock situation when they reach the obstacle avoidance distance because they are driving towards each other. Appropriate measures need to be taken to solve the deadlock of the two logistics robots to ensure the normal operation of the two logistics robots and complete the cargo handling task.

[0072] According to the priority division planning order, R2 with a low priority needs to take measures such as stopping and waiting or taking a detour to avoid collision. However, R2 will affect R1's arrival at the target point, so the planning order needs to be exchanged. R1 waits for 2s at node g2 and then drives after R2 passes the conflicting node to avoid collision. The planned time window is as follows Figure 5 As shown, Figure 5 (a) is the time window before extension, Figure 5 (b) is the extended time window.

[0073] If two logistics robots have the same priority and are at the same distance from the conflict grid g3, it is impossible to determine which robot will plan the path first, and the robot will fall into a deadlock state. In the case of planning failure, it is necessary to make a clearer division of the road rights of each logistics robot, and adjust the path planning order based on the optimal path of all logistics robots. Not only should the logistics robots be prioritized based on empty load, full load, etc., but the mutual occupation of paths between robots should also be considered. The path occupancy right of R1 is ρ1, and the path occupancy right of R2 is ρ2. The priority of the logistics robot is determined by the relationship between the path occupancy rights.

[0074] If the target point of R1 is on the optimal path of R2, then ρ1<ρ2 should be satisfied, and R2 has the right to priority path planning to avoid deadlock.

[0075] If the starting point of R1 is on the optimal path of R2, then ρ1>ρ2 should be satisfied, and R1 has the right to priority path planning to avoid deadlock.

[0076] According to the above two conditions, the present invention can perform road right analysis on all planned logistics robots and obtain the relationship between the priorities of each robot. In actual planning, high-priority logistics robots are given priority planning, and then low-priority logistics robots are optimized, which improves the success rate of planning to a certain extent.

[0077] When multiple logistics robots occupy each other with different priorities, the planning order cannot be obtained based solely on the priority level. When this happens, some logistics robots need to temporarily deviate from the optimal path and give way to other robots. After other robots pass, they return to the optimal path to continue their tasks. If there are many repeated sections of logistics robots that need to pass on the same path, and two robots with the same or different priorities have multiple time window conflicts, the low-priority logistics robot will re-plan the path and remove the path segments with path conflicts from the search list to avoid the problem of multiple logistics robots chasing each other on a path, causing the low-priority logistics robot to wait for the high-priority logistics robot for a long time.

[0078] 2. Time window conflict avoidance strategy

[0079] After completing the task allocation and path planning, the time window algorithm is used to detect the time window of the task path to determine whether there is a time window conflict between the paths. If there is no conflict, it can be executed according to the task path. If there is a conflict, the corresponding conflict avoidance strategy is adopted according to the corresponding conflict type to complete the time window arrangement and resolve the time window conflict. The conflict avoidance strategies of the time window algorithm generally include parking waiting, path replanning, and speed adjustment. The corresponding methods are three kinds of operation: translation, stretching, and compression of the time window. The time window translation operation refers to the low-priority robot stopping in front of the conflict node and waiting for the high-priority robot to pass first. The low-priority robot passes the node again, and the time window of the low-priority robot needs to be stretched and translated backward. If the two operations of time window stretching and compression involve the acceleration and deceleration of the robot, it will lead to an increase in the amount of calculation and the uncontrollable factors of the system.

[0080] The two measures commonly used in the avoidance strategy of time window conflicts are parking and waiting and path replanning. When multiple robots work at the same time, the path situation is relatively complicated. If only the parking and waiting strategy is adopted, it will cause blockages on each path, affecting the working efficiency of the logistics system. If the path replanning strategy is adopted, the handling task cannot be completed, because the time window conflict avoidance is started according to the task priority. The high-priority robots occupy too many paths, and the subsequent robots will fall into the problem of having no path to go. In order to solve the problem of deadlock and blockage of path conflict nodes, the present invention designs a conflict avoidance strategy that combines parking and waiting with path replanning, and the path replanning is divided into primary replanning and secondary replanning.

[0081] The path planning order of the logistics robot is obtained through priority division, and different conflict avoidance strategies are adopted according to different conflict types. If there is a node conflict, the path conflict avoidance can be completed by just stopping and waiting. For chasing conflicts and opposite conflicts, the low-priority robot needs to replan the path to solve the conflict problem, especially when the paths between robots are still occupied by each other, it is more necessary to analyze the right of way and select the appropriate avoidance strategy. The first-level path replanning means that when a time window conflict is detected, the conflicting nodes are removed from the search list, and the ant colony algorithm is used to replan the path from the starting point to the target point for the low-priority robot. If there is no conflict in the time window of the replanned path, it means that the replanning is successful. If there is still a time window conflict, it means that the replanning fails and it is necessary to wait for the position of other robots to change before replanning the path. The second-level path replanning means that by traversing the path information, the previous node of the conflicting node is found, and the path from the node to the target point is replanned for the low-priority robot. The time cost of the parking waiting and replanning measures is compared, and the shortest time solution is selected for execution.

[0082] 3. The multi-logistics robot path planning method designed by the present invention, based on the improved time window algorithm, orderly assigns the handling tasks to each logistics robot according to the handling task priority and the robot priority. High-priority handling tasks are assigned to high-priority robots, and low-priority handling tasks are assigned to low-priority logistics robots. Then, the ant colony algorithm is used to complete the path planning according to the starting coordinates and target coordinates of the handling tasks. The time window algorithm is then used to calculate the time windows of the logistics robots on each path segment, and the time windows are arranged in order on each path. Paths with time window conflicts are detected, and the path conflict types are analyzed. Then, corresponding measures are used to complete time window adjustments or path replanning until there are no time window conflicts on all paths. Figure 6 As shown, the path planning method specifically includes the following steps:

[0083] Step 1: Assign tasks to logistics robots in order according to their priority.

[0084] Step 2: Use the ant colony algorithm to complete path planning for the logistics robot and complete the time window calculation for each logistics robot.

[0085] Step 3: Determine whether the time windows of the current highest priority logistics robot overlap with the second highest priority logistics robot. If not, jump to step 7. If there is a time window overlap, execute the next step.

[0086] Step 4: Determine whether the path conflict type is a node conflict. If so, the low-priority logistics robot stops at the node before the conflict node and waits until the high-priority logistics robot passes the node and completes the time window update. Otherwise, execute the next step.

[0087] Step 5: Determine whether the path conflict type with time overlap is an opposite conflict. If it is an opposite conflict, calculate the parking waiting time of the low-priority logistics robot from the previous node entering the conflicting path, then remove the conflicting path from the search list, and re-plan the running time required for the path to the target point with this node as the starting point. Finally, select the plan with the shortest time consumption, update the time window of the plan and jump to step 3. If there is no opposite conflict, execute the next step.

[0088] Step 6: Determine whether the path conflict type with time overlap is a pursuit conflict. If it is a pursuit conflict, perform road right analysis on the two logistics robots. If the driving order of the two will not cause obstacles for the other party to reach the target point, jump to step 3; if it will cause obstacles, remove the conflicting path nodes of the two logistics robots from the search list, use the ant colony algorithm to search for a suboptimal path for the low-priority logistics robot again, complete the time window update, and jump to step 3.

[0089] Step 7: Determine whether all tasks are completed. If all tasks are completed, proceed to the next step. If not, jump to step 3.

[0090] Step 8: The algorithm ends.

[0091] 4. Experimental simulation and result analysis:

[0092] In order to verify whether the improved time window algorithm can solve the path conflict problem that occurs when multiple logistics robots are running in coordination in the logistics warehousing environment grid map. Firstly, the grid method is used to model the warehousing logistics environment, and then various path conflict types are simulated on MATLAB2016b to verify the feasibility of the improved algorithm. In the indoor warehousing working environment, there are usually cargo delivery points, charging stations, picking stations, and standby areas. The warehousing logistics operation environment grid map is as follows: Figure 7 As shown. Each grid coordinate is represented by the coordinate of the lower left corner of the grid; the positions of static obstacles such as delivery points and picking platforms in the work area are known, and dynamic obstacles are not considered when modeling the environment; in the grid map model, the logistics robot is regarded as a particle, and the coordinate information of its starting position and target position is obtained by the distribution of the handling task.

[0093] The grid map simulates the modeling of real logistics scenarios, and the division of each functional area in the map is consistent with that of a logistics warehouse. The entire grid map is a square with a length of 16m and a width of 16m. Four cargo picking platforms are set up for twelve logistics robots to complete cargo picking at the same time. Four charging stations can meet the charging needs of 8 logistics robots at the same time. There are 36 cargo delivery points. The size of the logistics robot is 60cm long, 60cm wide, and 30cm high. The stable driving speed is 1m / s. It can rotate 360° on the spot and is mainly responsible for completing the cargo handling work. The working time of each logistics robot at the delivery point and the picking platform is 5s to complete the delivery and pickup of goods. The picking platform is located on the upper and lower sides of the map, responsible for sorting the goods, and then the logistics robot transports them to the designated delivery point. The charging station is located in the waiting area, where idle logistics robots wait for task assignment. When the power of the logistics robot is lower than the set threshold, it stops accepting tasks and drives to the idle charging station for charging.

[0094] In order to facilitate the implementation of the simulation experiment, the logistics robot is set as a mass point in the map, the movement speed is 1m / s, the safety distance between logistics robots is one grid, the turning time in situ is Δt=0.5s, and 6 logistics robots are set to work simultaneously. The logistics robot set is R={R1, R2, R3, R4, R5, R6}, the task set is M={M1, M2, M3, M4, M5, M6}, and the task priority decreases as the task number increases. The 6 handling tasks are assigned to the 6 logistics robots, as shown in Table 3. The table describes in detail the task number, starting coordinates, target coordinates, task priority, and departure time corresponding to each logistics robot.

[0095] Table 3 Operation parameters of handling tasks

[0096]

[0097] According to the starting coordinates and target coordinates in Table 3, the improved ant colony algorithm is used to plan the paths of the six logistics robots. The path results of each logistics robot are shown in the following figure. Figure 8 shown.

[0098] From the path planning diagram, we can see that R1 and R2 have a chasing conflict and the target point is unreachable. R3 and R4 have a conflict in the opposite direction. The dotted path is the conflict path between the logistics robots. R5 and R6 do not have a path conflict and can directly perform the handling task according to the path. Subsequently, the time window calculation and path optimization are performed for the logistics robots with path conflicts. The time window calculation is performed for the handling task process performed by R1 and R2. The time window calculation results of R1 are shown in Table 4, and the time window calculation results of R2 are shown in Table 5.

[0099] Table 4 R1 grid dwell time

[0100]

[0101]

[0102] Table 5 R2 grid dwell time

[0103]

[0104] The time window diagram drawn according to the time window data in Table 4 and Table 5 is as follows: Fig. 9 As shown in the figure, the square mark is the node where the two logistics robots have a time window conflict. It can be seen that R1 and R2 have a time window conflict on the path (5,1)→(6,1). The path conflict type is a pursuit conflict, and R1 will also affect R2's arrival at the destination. R2 needs to wait until R1 completes the cargo picking and leaves the node before it can go to the destination.

[0105] Since the robot itself is equipped with obstacle avoidance sensors, there will be no collision during the forward and backward driving process. However, the target point R1 is at the previous coordinate of the target point R2, and the task priority of R1 is higher than that of R2. R2 needs to wait for R1 to complete the cargo picking and leave the path before continuing the task. The waiting time is 6s, which will cause congestion on the path. If the path replanning measure is taken, R2 does not need to stop and wait. The time required to perform the task according to the replanned path is shorter than the time required for the stop and wait measure. The schematic diagram of the path time window after replanning using the improved ant colony algorithm is shown in the figure below. Fig.10 As shown. Fig.10 It can be seen that after R2 replanned the path, there was no conflict in the time windows of the two logistics robots. The improved time window algorithm successfully solved the chasing conflict problem between the two logistics robots.

[0106] The time window data calculation is completed according to the path information of R3 and R4. The time window calculation results of R3 are shown in Table 6, and the time window calculation results of R4 are shown in Table 7. The starting coordinates of R3 are (6, 14), and the coordinates of the target point are (3, 10). The starting coordinates of R4 are (5, 14), and the coordinates of the target point are (8, 10).

[0107] Table 6 R3 grid dwell time

[0108]

[0109] Table 7 R4 grid dwell time

[0110]

[0111] According to the time window data in Table 6 and Table 7, the time window diagram is drawn, as shown in Fig.11 shown. Fig.11The square mark in the middle is the node where the two logistics robots have a time window conflict. R3 and R4 will collide with each other on the path (6,10)→(5,10). The two logistics robots will travel towards each other. Due to the infrared obstacle avoidance sensor, there will be no collision, but the two logistics robots will be deadlocked. If no pre-obstacle avoidance measures are taken, the path will be completely blocked and impassable. According to the task parameters, R3 has a higher task priority than R4. According to the conflict avoidance strategy, the low-priority logistics robot can choose to park and wait or re-plan the path to avoid the conflict. The time costs of the two solutions will be calculated separately, and the path solution with the shorter time cost will be selected for execution.

[0112] The priority of the transport task performed by R3 is higher than that of R4. If the stop-and-wait measure is adopted, R4 will stop and wait for 6 seconds at (5,11), waiting for R3 to pass the path (6,10)→(5,10), and then R4 will continue to drive to complete the transport task. If R4 takes a detour measure and replans the path to the target point from the node (5,11), it does not need to wait for 6 seconds, but only needs to wait for 1 second. After R3 passes the node (6,11), it drives according to the replanned path. The required task time is shorter than the time to stop and wait before executing the task. The schematic diagram of the time window for R4 to complete the path replanning using the improved ant colony algorithm is shown in the figure below. Fig.12 As shown. Fig.12 It can be seen that there will be no time window conflict between the two logistics robots that have completed the re-planning.

[0113] The improved time window algorithm of the present invention is used to detect the time windows of the four paths, and the conflict type of the paths with time window conflicts is analyzed, and then the corresponding avoidance strategy is adopted according to the path conflict type. The paths planned by the ant colony algorithm for the four logistics robots are replanned. The planned paths are as follows: Fig.13 As shown in the figure. Because the driving path of R2 is occupied by R1, the improved ant colony algorithm is used to re-plan the path for R2, and the obtained path does not conflict with R1. The path of R3 has not changed, and the path of R4 has been re-planned. It needs to wait for 1 second at the node (5,11) and then drive according to the re-planned path, as shown in the figure. Fig.13 As shown, R3 will continue to drive towards the target point after passing the conflict node. The path results of R5 and R6 also remain unchanged.

[0114] In summary, the local path planning algorithm based on the improved time window algorithm of the present invention can effectively solve the path conflict problem between multiple logistics robots in logistics scenarios, ensuring that multiple logistics robots can complete the predetermined handling tasks more safely and efficiently. The effectiveness of the improved time window algorithm of the present invention is verified through simulation experiments.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A multi-logistics robot path planning method based on an improved time window algorithm, characterized in that: The method allocates the handling tasks to each logistics robot in order according to the task priority and robot priority; allocates the high-priority handling tasks to the high-priority robots, and allocates the low-priority handling tasks to the low-priority logistics robots, and then uses the ant colony algorithm to complete the path planning according to the starting coordinates and target coordinates of the handling tasks, and then uses the improved time window algorithm to calculate the time window of the logistics robot on each path segment, arranges the time windows in order on each path, detects the paths with time window conflicts, and adjusts or replans the paths according to the time window conflict avoidance strategy until all paths have no time window conflicts; The method specifically comprises the following steps: S1: Assign tasks to logistics robots in order according to their priority; S2: Use the ant colony algorithm to complete the path planning for the logistics robot, and complete the time window calculation of each logistics robot according to the improved time window algorithm; S3: Determine whether the time windows of the current highest priority logistics robot and the second highest priority logistics robot overlap. If not, jump to step 7; if there is a time window overlap, execute step S4; S4: Determine whether the path conflict type with time overlap is a node conflict. If so, the low-priority logistics robot stops and waits at the node before the conflicting node until the high-priority logistics robot passes the node and completes the time window update; otherwise, execute step S5; S5: Determine whether the path conflict type with time overlap is an opposite conflict. If it is an opposite conflict, calculate the parking waiting time of the low-priority logistics robot from the previous node of the conflicting path, remove the conflicting path from the search list, and re-plan the running time required for the path to the target point with the node as the starting point. Finally, select the plan with the shortest time consumption, update the time window of the plan and jump to step S3; if there is no opposite conflict, execute step S6; S6: Determine whether the path conflict type with time overlap is a chasing conflict. If it is a chasing conflict, perform road right analysis on the two logistics robots. If the driving order of the two will not cause obstacles for each other to reach the target point, jump to step S3; if it will cause obstacles, remove the conflicting path nodes of the two logistics robots from the search list, use the ant colony algorithm to search for a suboptimal path for the low-priority logistics robot again, complete the time window update, and jump to step S3; S7: Determine whether all tasks are completed. If all tasks are completed, continue to step S8; if not, jump to step S3; S8: End.

2. The multi-logistics robot path planning method according to claim 1, characterized in that: In step S1, the rules for assigning task priorities are as follows: (1) handling tasks are assigned in descending order of priority; (2) each logistics robot can only be assigned one task at a time; (3) each handling task can only be assigned to one logistics robot.

3. The multi-logistics robot path planning method according to claim 1, characterized in that: In step S1, each logistics robot has a fixed ID number. The priority of the logistics robot is determined by the ID number. The robot with a higher priority will get the handling task first. The smaller the ID number, the higher the priority. The logistics robot status is divided into: picking status, running status, idle status, charging status, and fault status; logistics robots in charging status and fault status will not appear in the list of handling task assignments; logistics robots in picking status indicate that they are waiting for goods to be picked in the shelf area, and they do not exist in the sequence of task assignments; robots in running status indicate that they have completed task assignments and are going to the picking area or have completed picking and are going to the target area to deliver goods; robots in idle status will appear in the robot list of task assignments, and the list ranking order is arranged according to the size of the ID number; the ID number of the logistics robot in idle status determines the priority of the robot. The smaller the ID number, the higher the priority, and the handling task is assigned first.

4. The multi-logistics robot path planning method according to claim 1, characterized in that: The improved time window algorithm includes: after considering the status of the logistics robot, if the two logistics robots have the same priority, the priority of the logistics robot is determined by the relationship of the path occupancy right; If logistics robots R 1. The target point is the logistics robot R 2 On the optimal path, it should satisfy , R 2 has the right to prioritize path planning to avoid deadlock; among them, for R 1 path occupancy right, for R 2's path occupancy rights; If logistics robots R 1 Starting point in logistics robots R 2, then it should satisfy , Have the right to prioritize path planning to avoid deadlock.

5. The multi-logistics robot path planning method according to claim 4, characterized in that: The improved time window algorithm also includes: when multiple logistics robots occupy each other's priorities, different conflict avoidance strategies are adopted according to different conflict types.

Images